I'm trying to use FUBAR in place of the branch-site REL method in order to enjoy the computational speed-up; however, I do not see a straightforward way to return the branches and sites under selection in the output. I would like to estimate the proportion of sites under positive selection for each branch for 12,000 gene clusters. I'm able to run FUBAR just fine on our cluster non-interactively, it's just a matter of returning for each branch for each site if that site is under selection in the csv output file. Is it feasible for FUBAR to return this lineage-specific information?

Hi Jorge,

Unfortunately, there is no way of extracting branch-specific information from FUBAR, as the model explicitly assumes that omega is constant from branch to branch. We have some methods in the works for doing a FUBAR-style branch-site REL model, but those haven't been finished yet

Sergei